Proc. Natl. Acad. Sci. USAVol. 84, pp. 5316-5319, August 1987Developmental Biology

Spermatid-specific expression of protamine 1 in transgenic mice(germ cell/spermiogenesis/gene transfer)

JACQUES J. PESCHON*, RICHARD R. BEHRINGERt, RALPH L. BRINSTERt, AND RICHARD D. PALMITER**Department of Biochemistry and Howard Hughes Medical Institute, University of Washington, Seattle, WA 98195; and tLaboratory of ReproductivePhysiology, School of Veterinary Medicine, University of Pennsylvania, Philadelphia, PA 19104

Contributed by Ralph L. Brinster, May 6, 1987

ABSTRACT Protamines are abundant basic proteins in-volved in the condensation of sperm chromatin. In the mouse,protamine genes are transcribed postmeiotically in roundspermatids. We have cloned and sequenced the mouse prot-amine 1 gene. Ten lines of transgenic mice harboring markedprotamine 1 sequences were generated by microinjection offertilized eggs. Transcription of the transgene is restricted toround spermatids and in several cases exceeds that of theendogenous gene. The cis-acting sequences required for tissue-specific protamine expression reside on a 2.4-kilobase restric-tion fragment. Prospects for using transgenic mice to addressfundamental questions of male germ-cell development arediscussed.

Protamines are small, highly basic proteins that replacehistones and testis basic proteins during development ofmature spermatozoa (1). In mammals, extensive disulfidecrosslinking of protamine results in the formation of acompact chromatin structure devoid of transcriptional activ-ity. Two protamine variants have been identified in themouse, protamine 1 (mPl) and protamine 2 (mP2), differingin size, amino acid composition, and relative abundance (1,2). The functional significance of these differences is notknown.An analysis of protamine expression during prepubertal

development of the mouse testis (3), as well as in isolatedspermatogenic cell types (4, 5), indicates that mP1 and mP2are specifically transcribed in haploid, round spermatids. Theexistence of haploid gene expression has important biologicalramifications in that segregation and subsequent expressionof genes after meiosis could potentially result in phenotypicdifferences between individual spermatozoa. The extent towhich this occurs is not known, although it has beenpostulated as a mechanism explaining transmission-ratiodistortion, the non-Mendelian inheritance of specific alleles(6). A complication of this model is that spermatids developas a syncytium, interconnected by cytoplasmic bridges about1 um in diameter (6). This architecture is thought to assurethe synchronous development of clonally derived sperma-tids. Information transfer between spermatids could alsominimize or obviate the consequences of postmeiotic genetranscription; genetically haploid spermatids may be func-tionally diploid with respect to genes expressed after meiosis.The absence of distinguishing genetic markers specificallyexpressed in the haploid stages of spermatogenesis hasprecluded an investigation into the role of cytoplasmicbridges in maintaining an equal distribution of gene productsbetween individual spermatids.

Protamine synthesis is also regulated at the level oftranslation. Although mP1 mRNA is first detected in roundspermatids, it is stored in inactive ribonucleoprotein particlesfor up to 8 days until these cells differentiate into elongating

spermatids (7). Concomitant with the recruitment of mPlmRNA onto polysomes is extensive deadenylylation of thetranscript (7). Similar events are observed in the synthesis ofthe trout protamines, although these genes are active duringmeiosis (8).The isolation of a cDNA encoding mP1 has been described

(5). Using sequences derived from this cDNA, we cloned andcharacterized the mP1 gene. To develop a system in which toexamine the regulation and consequences of postmeioticgene expression in the mouse testis, we introduced markedcopies of this gene into mice by microinjection of fertilizedeggs. Expression of the transgene during male germ-celldevelopment, as well as in a variety of other tissues, wasmonitored.

MATERIALS AND METHODSCloning and Sequencing of the mPl Gene. CS7BL mouse

spleen DNA was partially digested with endonuclease Mbo Iand cloned into the BamHI site of X phage EMBL3. About300,000 recombinant phage were screened with a pair ofradiolabeled complementary oligonucleotides correspondingto nucleotides (nt) 232-287 of the mP1 cDNA (5). Recombi-nant phage DNA was isolated by standard procedures.Regions encompassing the mP1 gene were subcloned intoM13mp18 or 19 vectors (9) and sequenced by the chain-termination method (10). The sequence between the Nco Isite at nt -90 relative to the start of transcription and the BglII site at nt +625 was determined on both strands.Plasmid Construction and Microinjection. The sequence 5'

CAGTCTCAGGATCCACCATG 3' was inserted into a nu-clease Si-treated NcoI site at nt +95 by using complementaryoligonucleotides to differentiate between marked and endog-enous protamine transcripts. In addition, a 237-base-pair (bp)Bcl I-BamHI restriction fragment of simian virus 40 (SV40)was inserted into the unique Bgl II site at nt +625 to identifythe transgene. The resultant construct, mPl'-4.8, extendingfrom an Sst I site at -4.8 kilobases (kb) to the 3' Sal I site(Fig. 2), was separated from vector sequences and injectedinto the male pronucleus offertilized mouse eggs as described(11). mP1'-.88 was generated by isolating an Acc I-Sal Irestriction fragment from mPl'-4.8. Transgenic mice har-boring mP1'-4.8 or mP1'-.88 sequences were identified by"tail dot" analysis (12) using a nick-translated SV40 probe.Transgene copy number was determined by comparing thehybridization of wild-type and transgenic spleen DNA to anick-translated mP1 probe in a dot hybridization assay.RNA Extraction and Analysis. RNA used for primer exten-

sion and blot analyses was extracted by homogenization inguanidinium isothiocyanate, followed by precipitation withlithium chloride as described (13). Total nucleic acids ex-tracted from tissues by use of NaDodSO4 and proteinase K(14) were used for solution hybridization analysis.

Abbreviations: mPl and mP2, mouse protamines 1 and 2; SV40,simian virus 40; nt, nucleotide(s).

5316

The publication costs of this article were defrayed in part by page chargepayment. This article must therefore be hereby marked "advertisement"in accordance with 18 U.S.C. §1734 solely to indicate this fact.

Dow

nloa

ded

by g

uest

on

Janu

ary

4, 2

022

Proc. Natl. Acad. Sci. USA 84 (1987) 5317

For blot hybridization analysis, RNA was denatured in aformamide/formaldehyde buffer and then electrophoresed ina 1% agarose gel (15). Electrophoretically separated RNAswere blotted onto nitrocellulose and hybridized to the end-labeled oligonucleotide specific for the mPl'-4.8 andmPl'-.88 transcripts. Hybridization was performed at 420Cin a solution containing 6x SSPE (20x SSPE = 3.6 MNaCl/200 mM NaPj/20 mM EDTA, pH 7.4), 5x Denhardt'ssolution (lx = 0.02% Ficoll/0.02% polyvinylpyrrolidone/0.02% bovine serum albumin), 0.1% NaDodSO4, and 100 ,ugof denatured herring sperm DNA per ml. Blots were washedat 550C in Sx SSPE/0.1% NaDodSO4/0.1% NaPPj.

Primer-extension analysis was performed by hybridizingRNA to an end-labeled oligonucleotide primer complemen-tary to mP1 sequences between nt +99 and nt +119. Thisprimer hybridizes to both endogenous and transgene mP1transcripts but yields extension products that can be elec-trophoretically distinguished due to the oligonucleotide in-sertion at nt +95 in the transgene. Hybridizations werecarried out at 42°C for 90 min in 30 ,ul of660mM NaCl/22 mMTris Cl, pH 8/4.4 mM EDTA containing =20 fmol ofend-labeled primer and 2 ,g of testis RNA. Reaction mix-tures were then diluted to 80 ,ul with water. After precipita-tion with ethanol, nucleic acids were suspended in 30,l of100 mM KCl/100 mM Tris Cl, pH 8.3/10 mM MgCl2/10 mMdithiothreitol/500 ,uM each dNTP plus bovine serum albuminat 250 ,g/ml and 7 units ofavian myeloblastosis virus reversetranscriptase (Life Sciences, St. Petersburg, FL) and incu-bated at 42°C for 30 min. Reaction products were precipitatedwith ethanol, denatured in 50%o formamide at 100°C, andanalyzed by electrophoresis in an 8% acrylamide/8 M ureasequencing gel.

Solution hybridization analysis of total nucleic acid wascarried out as previously described (16), using the end-labeled oligonucleotide specific for mPl'-4.8 mRNA.

RESULTSCloning and Characterization of the mP1 Gene. Two com-

plementary oligonucleotides corresponding to a portion ofthe 3' untranslated region of the mP1 cDNA were used toscreen a C57BL mouse genomic library constructed in Xphage EMBL3. Approximately 300,000 recombinant phagewere analyzed and one positive clone was identified. South-ern blotting (17) experiments revealed that the mP1 gene issingle-copy (data not shown).Appropriate M13 subclones encompassing the mPl gene

were sequenced (Fig. 1). Two discrepancies between thissequence and that of the mP1 cDNA (5), located in the 3'untranslated region, are underlined. These nucleotide varia-tions are most likely due to strain polymorphisms betweenCD-1 and C57BL mice.The point of transcription initiation was identified by

primer extension using an oligonucleotide complementary tothe 5' untranslated region of mP1 (data not shown). Aconsensus "TATA" homology is present at nt -30 relativeto this point. The exon/intron boundaries were deduced fromthe mP1 cDNA sequence (5) and consensus splice signals(18). The 3' end of the mP1 transcript was determinedpreviously (5).

Expression of mP1 Following Gene Transfer into Mice. Todetermine if cloned sequences flanking the mP1 structuralgene could properly direct its expression to round spermatidsof the mouse testis, transgenic mice harboring these se-quences were generated. A 237-bp Bcl I-BamHI restrictionfragment of SV40 was cloned into a unique Bgl II sitedownstream of the mPl gene, and a pair of oligonucleotides(see Materials and Methods) was inserted into the Nco I siteat position +95. The resultant construct, mPl'-4.8 (Fig. 2),was used to create six lines of transgenic mice. Mice carrying

-560 gtctagtaat

-520 tctgcatcca

-480 atggggcctc

-440 ctctgagacc

-400 ggcgttggat

-360 gcattctctg

-320 acaaatccat

-280 ggttatctat

-240 acttaacacc

-200 ggatgccatc

-160 ctcctgatgc

-120 atttggacat

-80 aggtcctggt

-40 ctagtatcta

+1 acancccaca+41 cgacccaggt+81 caagccagca

+121 K SR S+11AAA.GCA.GGAG+11 R R R

+161 CAGACGGAGG

+201 aagtagaggg

+241 gggacttggg

gtccaacacc tccctcagtc caaacactgc

tgtggctccc atttatacct gaagcacttg

aatgttttac tagagcccac ccccctgcaa

ctctggattt gtctgtcagt gcctcactgg

aatttcttaa aaggtcaagt tccctcagca

agcagtctga agatgtgtgc tttcacagtt

gtggctgttt cacccacctg cctggccttg

caggacctag cctagaagca ggtgtgtggc

taagctgagt gactaactga acactcaagt

tttgtcactt cttgactgtg acacaagcaa

caaagccctg cccacccctc tcatgcccat

ggtacaggtc ctcactggc~catggtctgtgcctctttgac ttcataattc ctaggggcca

taagaggaag agggtgctgg ctcccaggcc

aaattccacc tgctcacagg ttggctggct

ggtgtcccctA R

gA GCCAG

R C RCAGATGCCGC

gctctgagcc agctcccggcY R C CR S

ATACCGATGC TGCCGCAGCA

R R R R R C RCGTCGCAGGC GAAGATGTCG

R R C C~R R R R R

AGGCGATGCT GCCGGCGGAG GAGGCGAAgt

ctgggctggg ctgtgggggg tgtggggtgc

catgtctggg agtccctctc accacttttc

+ 2 81 tt ctttctR C C R R R RSaY T28ttccttt agGATGCTGC CGTCGCCGCC GCTCATACAC

+321 CATAAGGTG

+361 ccatcaaaac+401 agcatctcgc

+441 aaaacaggag

+481 aatgttgaaa

+521 ggcttgggga

+561 ctggtcaaat

+601 gactaccaag

K K YAAAAALTACT AGatgcacag aatagcaagt

tcctgcgtga gaattttacc agacttcaag

cacatcttga aaaatgccac cgLccgatgacctgctaagg aacaatgcca cctgtcaataactcatccca ttcctgcctc ttggtccttg

ggggtgcgcg gatgtggtta gggaacatga

gggaagggct tcaaaagaat tcccaatatt

ccacctgtac aaatctEgiT

FIG. 1. Sequence of the mP1 gene. The TATA sequence, thepoint of transcription initiation, and the relevant restriction sites areunderlined. The two nucleotide differences between this sequenceand that of an mP1 cDNA clone are located in the 3' untranslatedregion and are underlined. The transcribed portion of the gene (nt1-495) is shown in bold type. Translated regions are shown inuppercase print. The mP1 amino acid sequence is indicated (standardone-letter amino acid symbols) above the exon sequences.

the transgene were identified using an SV40 probe, andexpression of this gene relative to the endogenous protamine1 gene was determined using a primer-extension protocol.The insertion of the oligonucleotides enabled us to electro-phoretically separate extension products derived fromendogenous and transgene mP1 transcripts by using a primer3' to the insertion. As shown in Fig. 3, mPl'-4.8 is expressedat variable levels in the adult testis of all six transgenic lines.

Blot hybridization analysis of RNA from various tissuesisolated from 6-week-old mice of line 1688-6 showed thatmPl'-4.8 expression in this line is restricted to the testis(Fig. 4). In addition, the same tissues of adult male mice fromlines 1567-5, 1688-4, and 1694-7 were analyzed for mPl'-4.8expression by a quantitative solution hybridization protocol(16). In each case, expression above background (back-ground was equivalent to 10 molecules of mRNA per cell)could only be detected in the testis, where the endogenousmP1 transcript is present at about 3000 molecules per cell(unpublished results).The time at which a transcript appears during the first wave

of spermatogenesis can be used to identify the spermatogeniccell type in which transcription initiates (1, 3). Expression ofmPl'-4.8 in line 1567-5 was monitored as a function ofneonatal development. As shown in Fig. 5, no expression ofthe transgene was observed in testis samples from 10- and

Developmental Biology: Peschon et al.

Dow

nloa

ded

by g

uest

on

Janu

ary

4, 2

022

5318 Developmental Biology: Peschon et al.

C)l

1688-6

0 0

z z

20mer 23

cc - w0m 00O 0

w LuI

37mpSV40

FIG. 2. Restriction map of mPl'-4.8. Relevant restriction sitesof the mP1 gene and their position relative to the point of transcrip-tion initiation are indicated. The SV40 and oligonucleotide insertionsused to identify the transgene and its transcript, respectively, aredepicted.

CD z~~LU LU-CO) CC X 4

wl > wJ 0 m

4.3Kb.

2.3 Kb-

19-day-old males. Expression was detected at day 25, how-ever. This pattern of expression parallels that of the endog-enous mP1 gene (3) and is indicative of round-spermatid-specific mPl'-4.8 transcription.To further localize the cis-acting sequences required for

efficient, tissue-specific mP1 transcription, expression ofmP1'-.88 was monitored in four transgenic males. In two ofthese males, levels of the marked transcript exceeded thoseofthe endogenous mP1 message (Fig. 6). As demonstrated formPl'-4.8, mPl'-.88 mRNA was not detected in othertissues (data not shown).

DISCUSSIONIn the mouse, several components of mature spermatozoa,such as the protamines, are encoded by genes transcribedduring the haploid stages of spermatogenesis (19). We havecloned and characterized the mP1 gene and found that it isspecifically expressed in round spermatids following genetransfer into mice. Using defined mP1 sequences, it should befeasible to target the expression of various gene products tomale germ cells in transgenic mice.cDNAs encoding protamine variants from several sources

have been described in detail (5, 20, 21). However, protaminegene structure has only been examined in the trout (22). Noneof the seven trout protamine genes contain introns (22). DNAsequence analysis of the mP1 gene revealed the presence ofa 90-bp intron. In addition, a considerable degree of sequence

I I I I I I10 in -t.0 4 ~~~~ Oh

in

in 4 4 4

.56 Kb._ _

FIG. 4. Tissue specificity of mPl'-4.8 expression. Total RNAsamples (10 ,ug) from various tissues of an adult male of the 1688-6line and from testis ofan adult wild-type (wt) male were examined byblot hybridization analysis using an oligonucleotide probe specific formPl'-4.8. The integrity of RNA in all lanes was established byhybridizing the blot to an actin probe (data not shown). Size markersat left indicate positions of various HindIII X restriction fragments.

identity 5' to the TATA homology is observed between mP1and the gene encoding the major protamine variant in mouse,mP2 (P. A. Johnson, J.J.P., P. C. Yelick, R.D.P., and N. B.Hecht, unpublished data). It is tempting to speculate thathomologies between nontranscribed portions of these tworelated genes reflect a functional significance of these se-

quences in transcriptional regulation.The mpl'-4.8 construct is expressed in the adult testis of

six independent transgenic mice (Fig. 3). In four cases,

1567-5

0 0) z J 3:V 0 10 4c

wt

I-m10

4.3 Kb-

2.3 Kb-

147b- 41:C_ mPl'-4.8

122b- mP-1110b-

90 b

FIG. 3. Expression of mPl'-4.8 relative to the endogenous mP1gene. Primer-extension analysis of adult testis RNA from wild-type(wt) and mPl'-4.8 transgenic mice was performed. Equal amountsof RNA were analyzed. The sizes (in bases, b) of various Msp IpBR322 restriction fragments used as size markers are shown. Thedifference in size between mPl'-4.8 and mP1 extension products isdue to the 20-bp oligonucleotide insertion in the transgene. The1688-4, 1694-7, and 1696-7 testis samples were from founder males.The remaining samples were from males of established lines.

FIG. 5. Expression ofmPl'-4.8 during testis development. Totaltestis RNA samples (10 Afig) from immature (10-, 19-, and 25-day-old)and adult (6- and 23-week-old) males of the 1567-5 line and from anadult wild-type (wt) mouse were examined by blot hybridizationanalysis using an oligonucleotide probe specific for mPl'-4.8. Theintegrity of RNA in all lanes was established by hybridizing the blotto an actin probe (data not shown).

0

-Ic

F- 200bp

wt

Lu -z -ahe_ uJY I-

.56Kb-

Proc. Natl. Acad. Sci. USA 84 (1987)

Dow

nloa

ded

by g

uest

on

Janu

ary

4, 2

022

Proc. Natl. Acad. Sci. USA 84 (1987) 5319

i@uGo co Ws ClI I I0 0) N T-

WT- o T-o 0 0

N:0 N4 N N~

147b- _ _mPl'-.88122b- go0 d mPl

110b-

90b-

FIG. 6. Expression of mPl'-.88 relative to the endogenous mP1gene. Primer-extension analysis of total adult testis RNA was

performed as described for Fig. 3. The 2009-6 and 2010-8 testissamples were from founder males. The remaining samples came frommales of established lines. wt, Wild type.

accumulation of the marked transcript exceeds that of theendogenous mPl message. This observation is most easilyexplained by the presence of multiple active copies of thetransgene. The endogenous mPl gene is single-copy, whereasthese four mice harbor between 4 and 20 copies of the markedgene (data not shown). The low level of expression observedin two mice could be a consequence of integration into anunfavorable chromosomal location. Alternatively, as pedi-grees from these two animals were not established, thepossibility exists that the founder males were germ-linemosaics (23), with only a fraction of the spermatids carryingthe transgene.One of our goals is to specifically direct gene products to

round spermatids by using mPl regulatory sequences. Thusit was necessary to establish not only that mPl'-4.8 wasefficiently transcribed in the adult testis but also that itsexpression was restricted to the appropriate cell type withinthis tissue. As shown in Fig. 5, mPl'-4.8 message was notpresent in day-19 testis but was present by day 25; hence weconclude that expression of this construct is restricted toround spermatids.These data suggest that the cis-acting sequence require-

ments for high-level tissue-specific protamine expression arefully contained within mPl'-4.8. To further delineate theserequirements, mPl'-.88 was tested. Although only two offour animals expressed this transgene in the adult testis,levels of the marked message exceeded those of the endog-enous mPl transcript (Fig. 6). Thus, a maximum of 2.4 kb ofDNA sequence, of which only 880 bp is 5' to the start oftranscription, is necessary for efficient protamine gene tran-scription. RNA blot analysis, similar to that shown in Fig. 4,showed that mPl'-.88 is specifically transcribed in the testis(data not shown). The reduced frequency of expression maybe a reflection of an increased sensitivity of the deletionconstruct to neighboring chromosomal sequences.As there are no continuous spermatid cell-culture systems,

one is restricted to transgenic mice to delineate the cis-actingrequirements for postmeiotic expression in the mouse testis.Preliminary results with an mPl-human growth hormonefusion gene suggest that sequences located 3' to the Bgl II siteat nt +625 are not required for efficient testis-specific mP1gene transcription.We have generated several lines of transgenic mice that

express levels of a marked mP1 mRNA exceeding those of

endogenous mPl. These males are fertile and properlytransmit the transgene, suggesting that increased levels ofmPl mRNA are not deleterious to spermatid maturation. It isformally possible that the oligonucleotide insertion into the 5'untranslated region of the transgene prevents translation.However, the heterogeneity in size of mPl'-4.8 mRNA(compare day 25 vs. adult, Fig. 5) suggests that some of thetranscripts are deadenylylated and are most likely enteringthe pool of total mPl mRNA available for translation.Perhaps posttranslational regulatory events operate to assurea precise ratio between mPl and mP2 deposited onto DNA.The two variants have considerably different amino acidcontents (1), paralleling those observed between the humanprotamine variants (24), and may have distinct roles in theDNA-packaging process. Alternatively, the presence of twoprotamine variants, at a specific ratio, may not be necessaryfor proper sperm function. In the rat and several othermammalian species, only one protamine variant, resemblingmPl in amino acid composition, is found in mature sperm (1).While the contribution of haploid gene expression to the

process of spermatogenesis in mouse is appreciated, its rolein dictating the identities and fates of genetically distinctspermatids is unclear (1, 6, 19). The ability to target geneproducts specifically to round spermatids in transgenic miceby use of mPl regulatory sequences will provide valuableinsights into both the regulation and consequences ofpostmeiotic gene expression in the mouse testis.

We thank our colleagues for critical evaluation of the manuscript.We also thank Thomas M. Wilkie for helpful discussions during thecourse of this work. Oligonucleotides were prepared by Dr. PatrickS. H. Chou and Dr. Yim Foon Lee of the Howard Hughes MedicalInstitute Chemical Synthesis Facility. This work was supported inpart by National Institutes of Health Grants HD09172 (R.D.P.) andHD17321 (R.L.B.) and by a Genetic System Corporation GraduateFellowship (J.J.P.).

1. Bellve, A. R. (1979) in Oxford Reviews of Reproductive Biology, ed.Finn, C. A. (Oxford Univ. Press, New York), Vol. 1, pp. 159-261.

2. Balhorn, R., Weston, S., Thomas, C. & Wyrobeck, A. J. (1984) Exp.Cell Res. 150, 298-308.

3. Hecht, N. B., Bower, P. A., Waters, S. H., Yelick, P. C. & Distel, R. J.(1986) Exp. Cell Res. 164, 183-190.

4. Kleene, K. C., Distel, R. J. & Hecht, N. B. (1983) Dev. Biol. 98,455-464.

5. Kleene, K. C., Distel, R. J. & Hecht, N. B. (1985) Biochemistry 24,719-722.

6. Erickson, R. P., Lewis, S. E. & Butley, M. (1981) J. Reprod. Immunol.3, 195-217.

7. Kleene, K. C., Distel, R. J. & Hecht, N. B. (1984) Dev. Biol. 105, 71-79.8. Sinclair, G. D. & Dixon, G. H. (1982) Biochemistry 21, 1869-1877.9. Norrander, J., Kempe, T. & Messing, J. (1983) Gene 26, 101-106.

10. Sanger, F., Nicklen, S. & Coulson, A. R. (1977) Proc. Natl. Acad. Sci.USA 74, 5463-5467.

11. Brinster, R. L., Chen, H. Y., Trumbauer, M. E., Yagle, M. K. &Palmiter, R. D. (1985) Proc. Natl. Acad. Sci. USA 82, 4438-4442.

12. Palmiter, R. D., Chen, H. Y. & Brinster, R. L. (1982) Cell 29, 701-710.13. Cathala, G., Savouret, J. F., Mendez, B. L., Karin, M., Martial, J. A. &

Baxter, J. D. (1983) DNA 2, 329-335.14. Durnam, D. M. & Palmiter, R. D. (1983) Anal. Biochem. 131, 385-393.15. Lehrach, H., Diamond, D., Wozney, J. M. & Boedtker, H. (1977)

Biochemistry 16, 4743-4751.16. Ornitz, D. M., Palmiter, R. D., Hammer, R. E., Brinster, R. L., Swift,

G. S. & MacDonald, R. J. (1985) Nature (London) 313, 600-603.17. Southern, E. M. (1975) J. Mol. Biol. 98, 503-517.18. Mount, S. M. (1982) Nucleic Acids Res. 10, 459-472.19. Hecht, N. B. (1986) in Experimental Approaches to Mammalian Embry-

onic Development, eds. Rossant, J. & Pedersen, R. (Cambridge Univ.Press, New York), pp. 151-193.

20. Krawetz, S. A., Connor, W. & Dixon, G. H. (1987) DNA 6, 47-57.21. States, J. C., Connor, W., Wosnick, M. A., Aiken, J. M., Gedamu, L.

& Dixon, G. H. (1982) Nucleic Acids Res. 10, 4551-4563.22. Aiken, J. M., McKenzie, D., Zhao, H.-Z., States, J. C. & Dixon, G. H.

(1983) Nucleic Acids Res. 11, 4907-4922.23. Wilkie, T. M., Brinster, R. L. & Palmiter, R. D. (1986) Dev. Biol. 118,

9-18.24. McKay, D. J., Renaux, B. S. & Dixon, G. H. (1986) Eur. J. Biochem.

156, 5-8.

Developmental Biology: Peschon et al.

Dow

nloa

ded

by g

uest

on

Janu

ary

4, 2

022